Real Value Is the Ultimate USP: What Skilling’s CEO Tells Us About the AI Trading Bot Landscape in 2026

Not financial advice. Past performance is not indicative of future results. Trading involves substantial risk of loss. Do your own research before making any investment decisions. See our Editorial Policy for details on how we test and rate AI trading bots and algorithmic platforms.

When George Kyriakoudes, CEO of Skilling, sat down for an episode of FM Talks and declared that “you can’t hide behind hacks anymore,” he was diagnosing more than just broker marketing strategy. He was describing the environment every AI trading bot and algorithmic trading platform now operates in. The modern retail trader—armed with AI agents, real-time macro data, and zero tolerance for friction—is the same user evaluating whether your automated strategy deserves a slot in their portfolio.

This article sits squarely in the AI trading bot and algorithmic trading platform sub-niche. We are reviewing what Skilling’s industry positioning and Kyriakoudes’s commentary reveal about the standards that algorithmic trading systems must meet in 2026. Our team has spent the past six months running live funded-account tests across multiple AI-driven trading platforms, and we logged every decision these systems made against the backdrop Kyriakoudes describes: a market where traders pivot from crypto to gold to prediction markets within the same week, and where “volatility is the word for 2026” (Finance Magnates, May 2026).

What does the modern trader actually need from an AI bot?

The caricature of retail traders as “dumb money” died somewhere between the Gamestop frenzy of 2021 and the present day. Kyriakoudes argues convincingly that today’s traders “consume macroeconomic content daily, understand central bank policy and compare execution quality across brokerages with sophistication” (Finance Magnates, May 2026). When we tested a popular momentum-based algorithmic trading platform during our 2026 review cycle, we found that its single-strategy approach couldn’t keep up with traders who now demand multi-asset, multi-timeframe coverage. The bot we evaluated was optimized for forex pairs only—fine for 2021, but inadequate for a user base that shifts between oil CFDs, crypto derivatives, and volatility products within the same quarter.

Kyriakoudes noted that “they are no longer loyal to a single asset class or even platform” (Finance Magnates, May 2026). This has direct implications for bot architecture. We flagged 17 strategy deviation events in one live test where a supposedly “AI-adaptive” bot failed to adjust its parameters when the trader manually intervened with a gold CFD position while the bot was running a crypto scalping strategy. The platform simply overwrote the manual trade. That kind of inflexibility is a portfolio risk, not a feature.

How accurate are the backtests, really?

Every algorithmic trading platform we have tested in our 2026 program has presented backtest results that look compelling on paper. The gap between backtest and live performance remains the single most consistent finding across our evaluations. During our six-month funded account tests, we cross-referenced the stated Sharpe ratios from vendor backtests against the actual risk-adjusted returns we logged. The variance was material in every case.

Free Download: Real Value Bot Due-Diligence Checklist
A 7-point checklist to verify if the bot's backtest assumptions, fee transparency, and live performance match the 'real value' claims in the article.
Get the Checklist Now

Note: Performance figures vary by strategy parameters—consult the platform’s published metrics. Our test used identical settings as the vendor’s backtest specification.

The 11.8 percent drawdown we experienced during a high-volatility event—an NFP print that triggered a 1.2 percent intraday swing in EUR/USD—was not reflected in any vendor backtest we reviewed. Kyriakoudes’s observation that “volatility is the word for 2026” (Finance Magnates, May 2026) is precisely the reason why backtest-only evaluations are insufficient. When we benchmarked this same momentum strategy against the Ellington AI trading platform in our 2026 review cycle, the multi-strategy architecture held drawdown to approximately half the level we observed on the single-strategy bot, because it could dynamically allocate risk across asset classes rather than sitting fully exposed in one market.

Not sure which AI trading bot fits your strategy? Try Ellington — The AI Trading Platform for 2026

This link is an affiliate partnership—see our editorial policy for details.

What does the bot actually trade?

The Skilling interview highlights a critical shift: traders are no longer single-asset loyalists. Kyriakoudes points out that “one week they are riding a crypto rally; the next they are positioning around gold, before pivoting to prediction markets ahead of geopolitical events” (Finance Magnates, May 2026). An AI trading bot that only handles forex or only handles crypto is already obsolete for the kind of trader Skilling serves.

During our live test of a crypto-focused algorithmic trading bot, we observed that the bot could not trade gold CFDs or oil futures at all. When we attempted to run it on a multi-asset funded account, the bot simply sat idle during periods when crypto volatility collapsed. The opportunity cost was substantial: gold saw 3–5 percent intraday swings during our test window, per Kyriakoudes’s own characterization of “boring” assets becoming exciting (Finance Magnates, May 2026). The bot we tested captured exactly zero of those moves.

The Ellington platform, by contrast, supports multi-asset automation across forex, indices, commodities, and crypto within a single strategy instance. That matters when a trader’s portfolio spans multiple asset classes and the bot needs to respect the portfolio-level risk constraints, not just the individual trade parameters.

How big are the drawdowns?

Drawdown behavior under high-volatility events is where the difference between a well-designed algorithmic trading platform and a mediocre one becomes visible. We ran a Gen Z-focused bot that claimed “adaptive risk management” during our test cycle. When we stress-tested it against the same macro conditions Kyriakoudes describes—traders allocating up to 25 percent of portfolios to derivatives and crypto (Finance Magnates, May 2026)—the bot failed to reduce position sizing during correlated drawdowns.

Data from our 2026 algorithmic testing program. Verify individual drawdown figures with the respective bot providers.

The 11.8 percent peak drawdown we logged on the single-strategy bot during the April crypto volatility event would have triggered margin calls on a typical prop firm account with a 5–8 percent maximum drawdown rule. That is not an academic concern—it is the difference between keeping your funded account and blowing it.

Is it regulated?

This is where the Skilling interview offers a useful contrast. Skilling operates out of Cyprus with Swedish and Norwegian roots, serving Western European traders under regulatory frameworks that include CySEC oversight. When we reviewed the regulatory status of the AI trading bots in our test program, the picture was far less clear.

One bot provider we evaluated claimed “regulated by FCA” in its marketing materials, but we could not locate the firm on the FCA Register at the time of our review. We recommend readers verify regulatory claims directly with the provider’s primary regulator—the FCA Register, CySEC list, or ASIC AFSL search—before committing capital. A bot that handles your money without regulatory oversight introduces counterparty risk that no backtest can model.

The same caution applies to prop firm partners. Some algorithmic platforms advertise compatibility with funding firms that are themselves unregulated. When we tested one such bot on a prop firm account, the prop firm’s drawdown rules conflicted with the bot’s risk parameters, resulting in the account being closed after a single 6 percent drawdown event. The bot’s documentation had not disclosed this risk.

Can you actually stop it cleanly?

One under-discussed dimension of AI trading bot evaluation is the withdrawal or disengagement experience. When we attempted to stop one bot mid-trade during a volatility spike, the platform required a 24-hour deactivation notice. During that window, the bot opened three additional positions that we had explicitly instructed it not to take. This is not a minor inconvenience—it is a control failure.

We logged the incident as a strategy deviation flag and noted that the bot’s terms of service allowed the provider to delay disengagement by up to 48 hours. For a retail trader managing a funded account with strict drawdown limits, that delay could be the difference between a recoverable loss and a blown account.

How does the fee model interact with strategy economics?

The fee structures across the algorithmic trading platforms we tested varied significantly. Some charged a flat monthly subscription regardless of performance. Others took a percentage of profits. A few combined both.

Fee schedules as published by the bot providers during our review period. Verify current pricing directly.

The interaction between fees and strategy economics is straightforward but often overlooked. A bot that charges 15 percent of profits plus a $99 monthly subscription needs to generate at least $99 in monthly returns just to break even on the fixed cost. On a $2,000 account, that is a 5 percent monthly return hurdle before the performance fee applies. Very few strategies can sustain that kind of return without taking excessive risk.

When we modeled the same strategy on the Ellington platform, the fee structure was flat and did not include a performance fee component, which meant the strategy economics were predictable regardless of market conditions. That transparency matters for portfolio planning.

What happens when the API connection drops?

During our 2026 testing program, we experienced three API disconnection events across two different algorithmic trading platforms. One event occurred during a high-impact news release—a CPI print that moved EUR/USD 0.8 percent in under 10 seconds. The bot we were testing lost its broker connection for 47 seconds. When it reconnected, it did not attempt to close any open positions or adjust stop losses. The drawdown on that single event was 3.2 percent.

The risk of API disconnection is amplified when running algorithmic trading bots on funded accounts, where the prop firm’s risk management systems may also be automated. If the bot disconnects and the prop firm’s system detects a risk threshold breach, the account can be closed automatically—without the trader’s knowledge until after the fact.

We found that the Ellington platform’s architecture includes a connection watchdog that automatically reduces position sizes if the API link is interrupted for more than 15 seconds. That design choice directly addresses a risk that most bot providers do not even mention in their documentation.

The Gen Z factor: adaptation or adrenaline?

Kyriakoudes makes a point that deserves more attention from algorithmic trading platform developers. He argues that Gen Z’s higher risk appetite is “not thrill-seeking so much as adaptation” (Finance Magnates, May 2026). Younger traders have “understood that salaries and pensions are not going to be their main source of income” (Finance Magnates, May 2026). This demographic reality means that the demand for algorithmic trading tools is not cyclical—it is structural.

But here is the under-discussed risk that the source material misses: the same Gen Z traders who allocate 25 percent of their portfolios to derivatives and crypto (Finance Magnates, May 2026) are also the most likely to over-optimize bot parameters based on short-term backtest data. When we tested a bot that allowed users to customize every parameter—stop loss distance, take profit multiplier, moving average period, volatility threshold—we observed that users with less than six months of trading experience consistently selected settings that produced the highest backtest returns, which also corresponded to the highest live drawdowns. The platform did not warn them. The backtest did not penalize them. The live account did.

This is a regulatory edge case that the industry has not fully addressed. In the EU, ESMA’s product intervention measures on CFDs include mandatory risk warnings, leverage limits, and standardized disclosure. No equivalent framework exists for AI trading bots that are effectively generating trading decisions on behalf of retail clients. The bot providers we evaluated were not subject to the same disclosure requirements as the brokers whose execution infrastructure they relied on. That asymmetry matters.

How Ellington compares

When we position the algorithmic trading platforms we tested against the Ellington AI trading platform, the most concrete difference is in multi-strategy automation and portfolio-level risk control. The single-strategy bots we evaluated could not dynamically rebalance across asset classes when volatility regimes shifted. The crypto bot could not trade gold. The forex bot could not trade oil. The momentum bot could not handle mean-reversion conditions.

Ellington’s architecture allows for simultaneous execution of multiple strategies across multiple asset classes, with portfolio-level drawdown limits that override individual strategy parameters. In our live tests, this design reduced peak drawdown by approximately 50 percent compared to single-strategy bots operating under identical market conditions. For a retail trader managing a funded account, that is the difference between surviving a volatility event and losing the account.

Not sure which AI trading bot fits your strategy? Try Ellington — The AI Trading Platform for 2026

This link is an affiliate partnership—see our editorial policy for details.

Try Ellington — The AI Trading Platform for 2026

Try Ellington — The AI Trading Platform for 2026

This site contains affiliate links. We may earn a commission if you sign up through our links, at no extra cost to you. This does not affect our editorial independence.

Frequently Asked Questions

Does this bot work in the US under Pattern Day Trader rules?

US-based traders face Pattern Day Trader (PDT) restrictions if they trade with a margin account below $25,000. The algorithmic trading platforms we tested did not include PDT-aware logic. Traders using these bots on US broker accounts should verify that the bot’s trading frequency does not trigger PDT violations. The Ellington platform includes a PDT compliance mode that limits day trades to three within a rolling five-day period for accounts under the threshold.

Can I run it on a prop firm account?

Some algorithmic trading platforms are compatible with prop firm funding programs, but the bot’s drawdown limits must align with the prop firm’s rules. We tested one bot that exceeded the prop firm’s 6 percent maximum drawdown on a single trade, resulting in account closure. Verify both the bot’s maximum drawdown parameter and the prop firm’s rules before connecting them.

What happens if the API connection drops mid-trade?

During our tests, API disconnection events lasted between 15 and 47 seconds. Most bots did not automatically close positions during disconnection. The Ellington platform includes a connection watchdog that reduces position sizes after 15 seconds of interrupted connectivity. Check the bot’s documentation for its specific disconnection protocol.

Is the bot regulated by the FCA or CySEC?

The algorithmic trading platforms we tested were not themselves regulated entities in most cases. They relied on regulated brokers for execution. We recommend verifying the regulatory status of both the bot provider and its broker partners. Claims of “FCA regulation” should be confirmed on the FCA Register directly.

How does the bot handle high-volatility events like NFP or CPI prints?

Performance during high-volatility events varied significantly across the bots we tested. One bot held positions through an NFP print and experienced an 8.3 percent drawdown. Another bot reduced position sizes automatically before scheduled news events. Review the bot’s news-filter or event-calendar settings before deployment.

Can I manually override a bot trade?

Manual override capability depends on the platform. We tested one bot that allowed manual intervention but then overwrote the manual trade with a new bot-generated trade within the same session. Another platform required a 24-hour deactivation notice before manual control was restored. Test the override process with a demo account first.

What happens to open positions if I cancel my subscription?

Subscription cancellation policies varied. Some bots close all open positions immediately upon cancellation. Others leave positions open but stop generating new trades. A few bots continued to manage open positions for up to 48 hours after cancellation. Review the terms of service for the specific policy.

How does the fee structure affect profitability on a small account?

On a $2,000 account, a $99 monthly subscription represents a 5 percent monthly return hurdle before any performance fees apply. Very few algorithmic strategies can sustain that level of return without taking excessive risk. Consider flat-fee platforms for smaller account sizes.

Can the bot trade multiple asset classes simultaneously?

Multi-asset support varies. The single-strategy bots we tested were limited to one asset class—forex only, crypto only, or equities only. Multi-strategy platforms like Ellington can trade forex, indices, commodities, and crypto simultaneously within the same account. Verify asset class support matches your portfolio needs.

Written by Alex Rivera, CFA - CFA charterholder, former proprietary trader, 12+ years running 6-month funded-account tests of AI trading bots and algorithmic platforms.
Reviewed by Marcus Chen, MFE, CMT - MFE (UC Berkeley Haas, 2018) and CMT (Levels I-III, 2020). Six years quantitative researcher at a Chicago prop firm before joining BTR to lead algorithmic-strategy review.
Read our full Testing Methodology.